--- title: "Resource library" description: "A curated library of AI learning resources — the courses, books, lectures, and repos that hold up, organized by topic and level." canonical_url: "https://fanout.sh/ai/resources" md_url: "https://fanout.sh/ai/resources.md" access: "public" --- # Resource library A curated library of AI learning resources — the courses, books, lectures, and repos that hold up, organized by topic and level. ## Books ### [Mathematics for Machine Learning](https://mml-book.github.io/) The foundational textbook connecting linear algebra, calculus, and probability to ML algorithms. Free PDF available. - Author: Deisenroht, Faisal & Ong - Tag: essential ### [Why Machines Learn](https://www.amazon.com/Why-Machines-Learn-Elegant-Behind/dp/0593185749) The elegant maths behind modern AI — a beautifully written exploration of the mathematical ideas powering deep learning. - Author: Anil Ananthaswamy - Tag: recommended ### [The Art of Doing Science and Engineering](https://www.amazon.com/Art-Doing-Science-Engineering-Learning/dp/1732265178) Learning to learn — Hamming's legendary guide to thinking clearly about science, engineering, and research methodology. - Author: Richard W. Hamming - Tag: mindset ### [Principles of Building AI Agents](/books/principles-of-building-ai-agents.pdf) A practical, no-hype guide to building AI agents — providers, models, prompts, tools, memory, and agentic workflows. Free PDF included. - Author: Sam Bhagwat - Tag: agents ### [Principles of Mathematics](https://www.amazon.com/Principles-Mathematics-Routledge-Classics-Bertrand/dp/0415487412) Russell's foundational work on mathematical logic — the philosophical underpinnings of mathematics that inform formal reasoning in AI. - Author: Bertrand Russell - Tag: foundation ### [Deep Learning](https://www.deeplearningbook.org/) The comprehensive deep learning textbook. Covers mathematical foundations, modern techniques, and research frontiers. Free online. - Author: Goodfellow, Bengio & Courville - Tag: essential ### [Pattern Recognition and Machine Learning](https://www.microsoft.com/en-us/research/publication/pattern-recognition-machine-learning/) A rigorous treatment of probabilistic ML — Bayesian methods, graphical models, and the mathematical theory behind learning. - Author: Christopher Bishop - Tag: advanced ### [Linear Algebra Done Right](https://linear.axler.net/) The best introduction to linear algebra for understanding vector spaces, eigenvalues, and the math behind neural networks. - Author: Sheldon Axler - Tag: foundation ### [Build a Large Language Model (From Scratch)](https://github.com/rasbt/LLMs-from-scratch) Step-by-step guide to building a ChatGPT-like LLM in PyTorch. Covers tokenization, pretraining, SFT, and RLHF. The companion GitHub repo has 90K+ stars. - Author: Sebastian Raschka - Tag: essential ### [Under the Hood](https://leanpub.com/under-the-hood) A 36-project workshop that builds and deliberately breaks every LLM layer — from scalar autograd and attention to training, KV cache, RLHF, quantization, multimodality, and specialist routing. - Author: Ramchand Kumaresan - Tag: llms ## Learning platforms ### [Math Academy](https://www.mathacademy.com/) AI-powered adaptive math learning platform. Covers everything from algebra to advanced linear algebra and calculus with spaced repetition. - Tag: adaptive ### [Khan Academy](https://www.khanacademy.org/math) Free, world-class education. Comprehensive math courses from arithmetic through multivariable calculus and linear algebra. - Tag: free ### [MIT OpenCourseWare](https://ocw.mit.edu/) Free access to MIT's entire math and CS curriculum. Includes 18.06 (Linear Algebra), 18.01/02 (Calculus), and 6.034 (AI). - Tag: free ### [Brilliant](https://brilliant.org/) Interactive problem-solving platform for math, science, and CS. Great for building intuition through hands-on exploration. - Tag: interactive ### [Notion — Study Tracker](https://www.notion.so/Math-Foundations-for-Machine-Learning-32174da94966803eb7e3c314e9e8d90b) Live study progress tracker — gets updated frequently as studies progress. Check current status and notes. - Tag: live ### [LLM Architecture Gallery](https://sebastianraschka.com/llm-architecture-gallery/) Visual gallery of 52+ LLM architectures by Sebastian Raschka — dense, MoE, hybrid, and recurrent models with clickable diagrams, fact sheets, and concept explainers. - Tag: interactive ### [TensorTonic](https://www.tensortonic.com/) LeetCode for ML papers — implement Transformers, BERT, ViT, DDPM, VAE, and GANs from scratch in cloud-native IDEs. 7,000 users in 2.5 months. - Tag: interactive ### [Dive into Deep Learning (d2l.ai)](https://d2l.ai/) Interactive textbook with runnable code in PyTorch, TensorFlow, and JAX. Used at 500+ universities for teaching deep learning. - Tag: free ### [AI by Hand (Prof. Tom Yeh)](https://www.byhand.ai/) Learn AI concepts by working through them by hand — a visual, intuitive approach that builds deep understanding before touching code. - Tag: interactive ### [Google Learn Your Way](https://learnyourway.withgoogle.com/) Uses LearnLM + Gemini to transform any PDF into 5 personalized learning formats. Students scored 78% vs 67% on retention tests. - Tag: adaptive ## Playlists ### [Essence of Linear Algebra](https://www.youtube.com/playlist?list=PLZHQObOWTQDPD3MizzM2xVFitgF8hE_ab) Visual, intuitive explanations of linear algebra concepts — vectors, matrices, eigenvalues, and more. The gold standard for building geometric intuition. - Channel: 3Blue1Brown - 16 videos - Tag: linear-algebra ### [Neural Networks](https://www.youtube.com/playlist?list=PLZHQObOWTQDNU6R1_67000Dx_ZCJB-3pi) Deep visual understanding of how neural networks actually work — gradient descent, backpropagation, and what neurons compute. - Channel: 3Blue1Brown - 4 videos - Tag: neural-nets ### [Essence of Calculus](https://www.youtube.com/playlist?list=PLZHQObOWTQDMsr9K-rj53DwVRMYO3t5Yr) Beautiful visual explanations of calculus from first principles — derivatives, integrals, and the chain rule. - Channel: 3Blue1Brown - 12 videos - Tag: calculus ### [MIT 18.06 Linear Algebra](https://www.youtube.com/playlist?list=PLE7DDD91010BC51F8) The legendary Gilbert Strang lectures on linear algebra. Complete MIT course — the definitive resource for deep understanding. - Channel: Gilbert Strang (MIT OCW) - 34 lectures - Tag: linear-algebra ### [StatQuest: Machine Learning](https://www.youtube.com/playlist?list=PLblh5JKOoLUICTaGLRoHQDuF_7q2GfuJF) Clear, no-nonsense explanations of ML algorithms and statistics. Perfect for building intuition before diving into math. - Channel: Josh Starmer - 100+ videos - Tag: ml-basics ### [Stanford CS229: Machine Learning](https://www.youtube.com/playlist?list=PLoROMvodv4rMiGQp3WXShtMGgzqpfVfbU) Andrew Ng's legendary ML course. Rigorous mathematical treatment of supervised learning, unsupervised learning, and deep learning. - Channel: Stanford Online - 20 lectures - Tag: ml-theory ### [Build a Large Language Model (From Scratch)](https://www.youtube.com/playlist?list=PLTKMiZHVd_2IIEsoJrWACkIxLRdfMlw11) Full video course implementing a ChatGPT-like LLM in PyTorch — tokenization, pretraining, SFT, RLHF, and inference. Companion to the 90K-star GitHub repo. - Channel: Sebastian Raschka - 20+ videos - Tag: neural-nets ### [Introduction to Deep Learning](https://sebastianraschka.com/blog/2021/dl-course.html) Complete university deep learning course — 170 video lectures with PyTorch code. Covers CNNs, RNNs, GANs, transformers, and more. - Channel: Sebastian Raschka - 170 lectures - Tag: neural-nets ### [Stanford CME 296: Diffusion & Large Vision Models](https://www.youtube.com/playlist?list=PLnFYHN9mfWzeggNsTxybm3vxfQpog43tJ) A rigorous course on diffusion, score matching, flow matching, latent spaces, guidance, modern image-generation architectures, and production training. - Channel: Afshine & Shervine Amidi (Stanford) - 6 lectures - Tag: generative-ai ### [Introduction to Machine Learning](https://sebastianraschka.com/blog/2021/ml-course.html) Full university ML course using scikit-learn. Python, tree methods, feature selection, and core ML algorithms. - Channel: Sebastian Raschka - 30+ lectures - Tag: ml-basics ## Channels ### [3Blue1Brown](https://www.youtube.com/@3blue1brown) The gold standard for visual math explanations. Linear algebra, calculus, neural networks — beautifully animated and deeply intuitive. - Handle: @3blue1brown - Subscribers: 6.5M - Tag: math ### [Andrej Karpathy](https://www.youtube.com/@AndrejKarpathy) Former Tesla AI director and OpenAI founding member. Neural Networks: Zero to Hero series builds GPT from scratch. - Handle: @AndrejKarpathy - Subscribers: 1M+ - Tag: deep-learning ### [Yannic Kilcher](https://www.youtube.com/@YannicKilcher) In-depth paper walkthroughs and ML news. One of the best channels for understanding cutting-edge research papers. - Handle: @YannicKilcher - Subscribers: 260K - Tag: papers ### [Two Minute Papers](https://www.youtube.com/@TwoMinutePapers) Bite-sized summaries of the latest AI research. Great for staying current on what's happening in the field. - Handle: @TwoMinutePapers - Subscribers: 1.5M - Tag: papers ### [StatQuest with Josh Starmer](https://www.youtube.com/@statquest) Crystal-clear explanations of statistics and ML algorithms. Makes complex concepts feel simple with humor and BAM! - Handle: @statquest - Subscribers: 1.2M - Tag: ml-basics ### [Sentdex](https://www.youtube.com/@sentdex) Practical Python and ML tutorials. Hands-on coding with real projects — from neural networks to self-driving cars. - Handle: @sentdex - Subscribers: 1.3M - Tag: coding ### [AI Explained](https://www.youtube.com/@aiexplained-official) Thoughtful analysis of AI developments, capabilities, and implications. Well-researched and balanced perspective. - Handle: @aiexplained-official - Subscribers: 500K - Tag: analysis ### [Umar Jamil](https://www.youtube.com/@uaborobot) Deep dives into transformer architectures, attention mechanisms, and LLM internals. Excellent for understanding how models work. - Handle: @uaborobot - Subscribers: 200K - Tag: deep-learning ### [Machine Learning Street Talk](https://www.youtube.com/@MachineLearningStreetTalk) Long-form interviews and discussions with top ML researchers. Deep technical conversations about the frontier of AI. - Handle: @MachineLearningStreetTalk - Subscribers: 180K - Tag: interviews ### [Sebastian Raschka](https://www.youtube.com/@SebastianRaschka) LLM research engineer. Build LLMs from scratch, architecture deep dives, LoRA/DoRA tutorials, and the LLM Architecture Gallery covering 52+ models. - Handle: @rasaborasbt - Subscribers: 100K+ - Tag: deep-learning --- This representation contains public Fanout content only. Protected Pro lessons, account data, billing, checkout, and pricing are not included. Browse the public content map: https://fanout.sh/sitemap.md